Companies remain excited by the potential of generative artificial intelligence (AI) to bring value to their enterprises and customers, but many are looking for the optimal path to pursue. To recall, generative AI is defined as artificial intelligence capable of generating text, images, video or other data using generative models, in response to a user’s instructions called prompts. Most recently, use cases have evolved to AI agents, a sophisticated tool capable of executing workflows autonomously and taking actions based on abstract goals defined by their users. Companies in all fields have experimented with different ways these emerging technologies can provide value to customers. This urgency is also partly owed to continued internal leadership and market pressures to release new generative AI-driven solutions. The main questions companies are asking: (1) where should be the focus?; and (2) what is the best path forward?

This is a generative AI playbook that provides companies a framework on how to prioritize use cases and start building. The key elements covered are:
I. Developing solid use cases
II. Developing the Solution: From Model Selection to Architecture
III. Assessing Model Use and Determining the Defensive Moats
IV. Risk Mitigation and Evaluation to Enable Enterprise Readiness for Generative AI
V. Why UI/UX and System Design Will Define AI for the Knowledge Worker
VI. Conclusion

This whitepaper is a refresh and updated perspective from our first. We view that given exponential advancements in these technologies, it was imperative for us to reframe our initial thinking. Some sections involve only modest changes, while others are rather substantive rewrites. While the focus of this paper will largely reflect on generative AI solutions that benefit external customers, our observations could also be extended to solutions that benefit employees and improve internal company processes.